SAP NetWeaver Business Warehouse
Page tree
Skip to end of metadata
Go to start of metadata

Overview

Aggregates allow you to improve the performance of BEx queries when data is read from an InfoCube. An aggregate is a materialized view on a subset of an InfoCube. It can contain characteristics and navigational attributes and can be restricted to fixed values (F) and hierarchy levels (H). Its purpose is to reduce the amount of data read from the DB during query execution.

A BEx query always tries to use performance-optimized sources by checking the sources from which it can draw the requested data. It checks them in the following order:

  1. OLAP cache

  2. BW Accelerator index

  3. Relational aggregates from the database (not supported when SAP HANA is used )

  4. InfoCubes from the database

So, if there is no cache entry or BWA index available, the system takes proper aggregates in case some exist. When the data in the cube changes, then the aggregates need to get adjusted. The same must be done when there are changes to the master data of affected characteristics (navigation attributes!) and hierarchies. Hence we have the following important processes:

  • Activation&Filling
  • RollUp - new transactional data (requests) is transferred to the aggregates
  • ChangeRun - aggregates are adjusted when there are relevant changes of master data and hierarchies
CSS component: BW-BEX-OT-AGGR

Aggregates used in Queries

When the query is executed, the system checks whether there are proper aggregates (if there is no cache entry or BWA index) available. If yes, the aggregates are used instead of the basis cube. The most convenient way to check whether data is being read from an aggregate, a BW Accelerator Index or an InfoCube is to run the query in the Query Monitor (transaction RSRT) by using a corresponding debug option. There you can also temporarily deactivate the usage of one or all aggregates in order to test them for performance purposes or to analyze data consistency.

For further details check the following example

When is an Aggregate used by a Query ?

Error Analysis

In case a query (or LISTCUBE which can use aggregates as well), which reads data from aggregates, displays incorrect data, it is recommended to proceed as described in the following note

1174503   Analysis for incorrect results using aggregates

If you want to check the consistency of aggregates on a regular basis you can use the transaction RSDDAGGRCHECK, for details please review the SAP Online Documentation

Performing Checks for Aggregates

Transactions and Tables

  • Transactions
    • RSDDV
    • RSDDAGGRCHECK 
    • LISTCUBE which can use aggregates as well
  • Tables
    • RSDDAGGRDIR
    • RSDDAGGR_V
    • RSDDAGGRCOMP

SAP Support Troubleshooting Guideline  

 SSG BW-BEX-OT-AGGR

SAP Online Documentation 

Performance Optimization with Aggregates

SAP Consulting Notes 

  • 903886    Hierarchy and attribute changerun (BW3x Release)
  • 1388570   BW Change Run (BW7x Release)
  • 1174503   Analysis for incorrect results using aggregates
  • 1053605   Termination 'GET_STATE_OF_AGGREGATES_CR-2-' in change run  
  • 1983809   Changerun aborts with GET_STATE_OF_AGGREGATES_CR-1-
  • 1907954   InfoCube or Attributes are Locked by Terminated Change Run.
  • 854092     During initial aggregate fill a COMMIT is eventually needed
  • 1085914   Restarting a change run results in error RSDD 85 
  • 1117724   Change run and condensing (BW 7.X) 
  • 1326211   BW change run: Repairing a terminated predecessor 
  • 534630    Parallel processing of Change Run 
  • 590370    Too many uncompressed request (f table partitions) 
  • 825927    The BW Changerun: CR_MAXWAIT 
  • 903886    Hierarchy and attribute change run 
  • 994005   Poor query performance when using aggregate 

Link to this page: https://wiki.scn.sap.com/wiki/x/hwBnFQ